|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
24.27.41.140
Tube sheets show typical operating conditions for power tubes, but I have never achieved those numbers in practice. In some cases, the results were pretty far off, maybe only 70% of the output power numbers shown by tube manufacturers. The two examples below represent amplifiers I built recently that produced considerably less than the sheets indicate.
The only explanation I have for this is loss through the output transformers. However, I wouldn't have expected it to be so significant. Is there a typical number for transformer loss in this respect? And all else being equal (impedance and power rating), are smaller, less expensive transformers less efficient, even at midrange? I'm just trying to establish a guide for this, something to help predict real world results at the beginning of a project.
--------------------------
Buy Chinese. Bury freedom.
Edits: 09/15/21Follow Ups:
I think it's time for Deathtube to spend some time at Camp LowMu.He does not contribute at all to these technical discussions but rather clogs them with his trolling and shilling.
I thought any sales or marketing was prohibited on this forum? Just about every post from him is shilling Serious Stereo.
But then I do see a loophole here. Serious Stereo has been out of business way before COVID so don't try to use that excuse. No WEB page, the URL is up for sale.
I guess ity's OK to shill for a dead company and product.
Edits: 09/22/21
"I think it's time for Deathtube to spend some time at Camp LowMu."Power loss can be directly attributed to poor system wiring, not only losses found in substandard OPT's.
That suggestion was met with overt hostility by the usual AA goon squad.
Since you are NOT a moderator, your opinion does not really matter regarding DT667's continued participation on AA forums.
But thanks for sharing, sir.
Have a nice evening.
DT667
Edits: 09/22/21
After scouring various RCA literature and searching all over the Web for examples of comparable 6L6-type amps, I think it's possible to infer a few relevant points. First, I believe the output power data in tube manuals is derived mathematically from the curves and load lines, not from physical testing. This means not only that no real world losses are included, but also that there's probably more room for error in the spec than would otherwise be the case. This assumes, of course, that the numbers are legitimate and that engineers of the day weren't influenced by the competitive nature of profitability. Second, the real output power from this amp is in agreement with others. The examples I found that make significantly more power all utilize much higher voltages (such as found in Fender guitar amps) or parallel output tubes. In fact, this amp seems to be performing as well as any, given its loading and voltage levels.
The other component clouding this issue is the original manufacturer's rating of 35W per channel. I never tested the amp for output power in its original configuration. A full restoration would have been required, and that wouldn't have been particularly useful once the transformers were transferred to the new chassis with a different topology. This admittedly opens areas of concern, but everything possible has been done to ensure the new design makes maximum use of the transformers. Rather, I believe the original rating was probably determined on the basis of "music power," and as such, would not have accounted for the power supply sag that occurs during continuous sine wave testing. I have a different 6L6 model from the same manufacturer that utilizes lower supply voltages and is rated 30W per channel. Based on the voltages, I suspect it only actually produces 25W or so. I'll test it in the near future, and if my suspicions are correct, the original power ratings will be rendered mostly meaningless. We'll see.
--------------------------
Buy Chinese. Bury freedom.
Won't the RMS voltage at one anode times the average DC current through the cathode resistor (minus screen current) for that tube equal 1/2 the power being delivered to the transformer? Or will power factor or other issues get in the way of this simple measurement? There's not much phase shift through the transformer at these frequencies, so I wouldn't think its reactance would create a problem.EDIT: The current measurement would need to be AC signal current, not average DC. In fact, it might be necessary to insert a low-value resistor in series with the anode temporarily and calculate anode current by measuring the AC voltage drop across that. Otherwise, signal current into the screen would skew the result.
--------------------------
Buy Chinese. Bury freedom.
Edits: 09/17/21
I made a measurement with the amplifier running at 60 Hz (a frequency that ensures the accuracy of my DVM). With 25V RMS on one anode (which would be 50V anode-to-anode), the loaded 8 ohm tap of the output transformer produces 1.77V RMS. Based on the 6.6K:8 impedance ratio of the transformer (give or take), the secondary should produce 1.74V RMS. This would certainly appear to indicate that the transformer is not the culprit.
--------------------------
Buy Chinese. Bury freedom.
TK,
Remember when they did these numbers they are driving the tubes direct coupled and therefore overcome the slew rate current and the grid current draw that happens capacitor coupled.
I have used the MQ tapped choke and gotten real close to the numbers for 6V6's in my guitar amps without issue. The low dcr and direct drive I used made it real close.
Damping also has an effect even with feedback which by equations lowers the output but in this case rarely has an effect on overall power.
Thanks,
Gordon
J. Gordon Rankin
Gordon, there's no slew rate limiting at the midrange frequencies where these measurements are being made. Pentodes don't exhibit sufficient input capacitance fro that to occur, and the empirical proof is that open loop gain and output power show no significant change with frequency until the input sweeps at least an order of magnitude higher. If you see greater output with direct coupling, it can only mean you're pushing into AB2 and drawing grid current. The charts I posted are specifically noted as being AB1 with no grid current and no driving power required.
As for FB, every amplifier I've ever built produced slightly more power (at the threshold of clipping) with feedback applied. That's due to the corrective nature of the loop.
--------------------------
Buy Chinese. Bury freedom.
Better use an interstage transformer to drive the end tubes instead of a grid choke.
You are losing power everywhere the wiring and connections between amp and speakers/crossover networks have not been optimized.
This is not just a problem that happens in low quality output transformers.
DT667
Did RCA or Sylvania "optimize the wiring"?
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
"Did RCA or Sylvania "optimize the wiring?"
As it pertains to what, OPT's or internal amplifier wiring?
Doubtful. Those finer aspects of the craft are rarely, if ever design considerations.
The standard engineer works to a price point given to them by finance or executive leadership who want the quickest path to revenue.
Innovation is now driven by disruption in a supply chain caused by global pandemic.
Getting raw materials to create any product is more an ordeal than ever.
DT667
"As it pertains to what"
As it pertains to the rated output wattage that TK is comparing against.
When the tube makers did their tests and wrote the data sheets (that TK is looking at and comparing his results to theirs) did they make those tests in a circuit where the wiring was optimized to your standards?
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
"When the tube makers did their tests and wrote the data sheets (that TK is looking at and comparing his results to theirs) did they make those tests in a circuit where the wiring was optimized to your standards?"
The only wiring standards that really matter are those that will elevate an amplifier to the "trade show" level or beyond.
DT667
nothing on the marked influence of the room in question.
Room
Speaker
Preamp
amp
In that order.
The source is another topic altogether, though quite important.
The Mind has No Firewall~ U.S. Army War College.
Ideally, musician/environment and speaker/room should sound as close to the same as possible. How often does that happen?
Lots of things in between to "color" the sound.
This is just for discussion, guys.At the trade shows I participated in,
every once in a while, a trade expert would
come in and listen, and then keep coming
back and listening over several days of
the show. The person in question is narrowing
the trade show rooms down to a very few good ones
on the last day of the show.One gentleman who was doing this was a
Hollywood sound engineer. He finally commented
on the last day of the show.Here is what he said-- paraphrased, but
accurate anyway--:"I understand what this room and its
components are doing. The sources are
excellent and are neutral. The amps are
perfect-- just what should happen.The speaker fit to the room is OK, I think,
but I hear a slight lower midrange depression.
This is what I always get when using ALTEC-based
transducers, including GPA, so it's what I expect.I'm buying everything in this room, so start packing
it up, guys.I learned from him and others-- over time--
that professional listeners can tell
exactly what's going in all parts of a music
system any time they want to.-Dennis-
Edits: 09/18/21 09/18/21 09/18/21 09/18/21
"This is just for discussion, guys"No, not this, not any discussion worth.
Dennis, i think (hope) you can do better then what you wrote before
Edits: 09/19/21
and we are clearly not worthy.
Or we could properly conclude that you are on dog food, and being the best salesman you know how to be.
I am going to consistently bet on the latter.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
"and we are clearly not worthy."
True enough, Pak.
Keep in mind that DF is a trade show veteran and professional tube amp builder, unlike yourself or most of us that participate on AA forums.
You can't realistically expect to reach DF's level in the audiophile game.
We can improve our systems via experimenting with the "concepts" that have been generously shared with us by DF and the other MLP builders.
That will result in domination of personal listening rooms and local audio club meetings.
DT667
"You can't realistically expect to reach DF's level in the audiophile game. "
What most of us are doing already exceeds what Dennis does.
Most of us use a power supply filter that works, Dennis doesn't.
Most of us use a driver stage that has a low enough output impedance and a high enough current delivery capability to properly drive the Miller capacitance of the output tube, Dennis doesn't.
Most of us use a cathode bypass capacitor that is sized to insure that there is no local current feedback all the way to the lowest frequency of interest so the damping factor of the amplifier is not compromised. Dennis doesn't.
I don't know where you got the idea that Dennis is some kind of special, he's not. The fact that you think he is just shows that you don't understand audio electronics very well.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Not using cathode bias resistor bypass caps.
Fixed bias. Decent filter design, and low output Z from the pots and where possible, grid chokes. (so far it has been extremely possible)
The rest ought to be obvious...DT and DF are building tone boxes for the folk susceptible to the particular brand of Marketing they are best skilled at throwing about. I guess for them it is a game; how much longer can we appear to get away with it, and can we direct the narrative so as to 'elevate' the BS?
No thanks, not interested...those rules bite, the reason for playing is dishonest, and you can't make me think playing it is useful...LOL
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
"I guess for them it is a game; how much longer can we appear to get away with it, and can we direct the narrative so as to 'elevate' the BS?"
For DF, it has been "serious business" as a profession audio products manufacturer operating at the highest levels of the industry and creating legendary pieces of high fidelity SE tube amplifiers - the best in class.
"No thanks, not interested...those rules bite, the reason for playing is dishonest, and you can't make me think playing it is useful...LOL
cheers."
Pak, if you are content with Class A PP tubes and mentoring a small group of people in a local audio club or policing audio forums, then fine.
But with ZERO profession trade show or professional audio industry experience, neither yourself or the majority of DF critics are at his level in the game. Not now , not ever.
The goose flock has been roasted again.
DT667
Gosh! I thought you were much better
informed than this!
-Dennis-
I am.Your driver stage can not properly drive the Miller capacitance of the output tube. It's output impedance is too high and it can not deliver enough current.
The total value of the capacitance across the 2a3 cathode resistor is too small to prevent local current feedback as the lower frequencies.
Your power supply filter is just a mess.
Prove me wrong with hard, technical facts. blah blah blah won't do this time.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 09/20/21
Maybe if he adds a 4th 'blah'?
On a serious note, it is all just effects box building. Interesting at first, and useful at shows where something technicolour is desired. Living with such things is always less than ideal.
Turn it on, listen...and listen...and listen...and listen...and not feel the need to do anything but feed it more music. This includes instances where you feel called upon to 'turn it UP!' a bit...LOL
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
"What most of us are doing already exceeds what Dennis does."
That is one of the most absurd statements ever posted on Tube DIY.
Where are the trade show reports or industry reviews? Nothing to see.
When DT667 encounters geese flocks at the park, they certainly enjoy honking and hissing, but never pose a serious threat.
Same thing here.
DT667
You have got to understand where your
detractors are coming from.
Most of them are into fixed-bias (unreliable),
1950's style tube amplification.
They believe all the old tube charts, even
though they have a really hard time getting
all the power the tube mfr's hinted at,
but never proved in the amp owner's
equipment.
A very few are into higher power SE tube
amps, but never realized that conventional
power supply technology is dead-wrong for
them. It was designed for P/P amps.
People still ignore the extreme importance
of mounting components very close to the usage
point-- something that doesn't make nice
pictures, but can be built rugged AND have
a wide musical bandwidth as well.
But it doesn't look like a clean P/P amp
layout-- and it can't if it's to compete
with the best amplifiers out there.
Low-DCR power sources are a make-or-break
item for a good SE amp. That's because only
one output tube is doing the signal sampling
for the speaker. No, such a P.S. design
won't satisfy people who try to splice
push/pull power supplies onto a SE amp.
The results are always dismal.
In a small SE amp, driver grid admittance
is of extreme importance. The higher you run the driver
plate current, the more of a "tuner" you are building.
The key here is to LISTEN to what is happening and
work with different values until you get it right,
commensurate with what the tube can run for a long
time with no problems.
Very few tubes that can actually perform well under
this requirement were ever made. The ones that do work
deliver stunning results.
There's always more, but just don't let them get
you down, they're selling tomatoes-- you might like Peaches.
-Dennis-
.
--
"You have got to understand where your
detractors are coming from."
The same mindset as Stalin , Mao and Dennis Rodman's DPRK buddies
DT667
You both seem to like sour grapes.
I have been there, done that, and not like to do that again
Do i hear a trial & error approach?If you have not the knowledge that's the only thing you can do., listen, listen listen ........
Edits: 09/20/21
Trial and error have nothing whatsoever to
do with good design.
Good design starts with conservative operating
points that assure that tubes and other parts will last
for years AND NOT CHANGE..
Once those parameters are SET FIRST, then theory and
listening can take place== AFTER reliability has BEEN
GUARANTEED by applying conservative design know-how.
Parts are then found that will deliver outstanding
musical performance WITH THE CONSERVATIVE OPERATING
VALUES THAT ARE SET INTO PLACE FIRST.... before anything
else is done or considered.
The challenge here is to find ways to obtain stellar
performance from ultra-conservative design, which
IS NECESSARY to have good equipment.
I have spent many years using commercial and DIY
tube gear. I never liked any of it.
Why? Performance degrades as tubes age. Reliability
suffers from power boosting crapola such as fixed
bias. Everyone wants high power from tubes.
Sonically, this is simply stupid because TUBES
DON'T LIKE THAT-- even if their plate curves
suggest otherwise. This is just one of several
areas where conventional tube equipment is mostly junk.
Along came solid-state. Once again, engineers ran
parameters based on theory and idealistic demands
which engineers expected the devices to obey. Once
again, the devices DID NOT LIKE what the engineers
were doing with them.
As a result pf this approach, it took almost 50 years
for solid state to finally reproduce music in a
convincing manner. During the last 8 years or so,
solid state equipment has finally achieved a first-ever,
High Fidelity relationship-- to music, not just to theory.
If any of you run autos in racing, or run off-road
equipment, you have spent a lot of money on computer
and SCR failures-- fixing them.
That was another standard, theory-based approach. It
also took 50 years to get a decent automobile computer.
Tube designers don't have to build junk. NO adjustments
are ever needed in truly good equipment. Tubes, in good
equipment, don't get softer and softer as the tubes age.
That is-- not when they are operated right-- and that
seldom agrees-- if ever- with published theory and assumed
correct operating points. Show me a design that sounds and
performs exactly the same after 20 years of use. Day in,
day out, with no tube replacements-- ever, unless one was
defective to start with.
Tubes still sound better-- more lifelike, more real.
You haven't lived until your amps perform like a good
furnace or refrigerator-- never needing anything.......
THAT is what makes your life worthwhile-- worth having
the musical equipment in the home.... when it never calls
attention to itself, but always plays music just like
it did last year, and the year before that, and the
year before that...... and.......on and on.
It's easy to criticize an approach to something when
you don't understand WHY the owner wanted it in the first
place... but you CAN understand-- if you really care enough.
Music is a lifestyle for some people. Building amps is
a totally different lifestyle for tinkerers and builders---
To please amp builders isn't what the amp is supposed
to do. It's best purpose is to play music flawlessly over
many years with NO attention required from its owner..
-Dennis-
What a fine salesman you are denis. I am at a loss to express my admiration for such a skill.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
"Why? Performance degrades as tubes age. Reliability
suffers from power boosting crapola such as fixed
bias. Everyone wants high power from tubes."
Pak like to "red plate" his PP tube finals.
Sounds a lot like the torture tests that Tube Lab does to sweep tubes.
That practice is probably OK for basement or local audio club events.
DT667
I go well beyond red there DT... :) If you knew anything about it, you'd say I was being too conservative and should run them hotter.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
"I go well beyond red there DT... :) If you knew anything about it, you'd say I was being too conservative and should run them hotter."
The only reason to use red plating transmitting tube finals is if you have inefficient transducers.
There are enough 3C24 and Eimac videos on YT of people doing this sort of thing.
Why do that when great audio tubes like JJ2A3-40 are available?
To each his own, Pak.
If toxic jabs don't get you, 2KV+ anodes will eventually.
DT667
well...once again you prove your ignorance beyond question.
As for 2kV, another area you insist on displaying your substantial ignorance.
As far as transducer efficiency, I don't think I exceed your gold standard by anything less than 6 dB at the 1W level of input.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
"As far as transducer efficiency, I don't think I exceed your gold standard by anything less than 6 dB at the 1W level of input."
If the first watt sucks, what is the point of adding hundreds of red plating PP sweep tube watts, Pak?
Call DF and have him explain the merits of JJ 2A3-40 as used in a high quality SE tube amplifier circuit.
Have a nice evening, sir.
DT667
Nice dodge there dead-toob. Keep right on going, at some point you may come upon something you can top somebody at.Also, just to be clear, I have no red-plating tubes. None, zero. Is that clear enough... :) I certainly have no red-plating sweep tubes. Those I run only a bit less conservatively than you suggest.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
Edits: 09/21/21
"Nice dodge there dead-toob. Keep right on going, at some point you may come upon something you can top somebody at."
Bobbing and weaving like a audio ninja, Pak.
Always keeping them off balance.
DT667 is already building SE DC 12W class circuits that are better than 95% of what is out there.
The other 5% is the MLP level of "God-like" performance , that will take more R&D time and $$$ to attain.
Enjoy your red plate PP tube amps and +6dB higher efficiency that DF's GPA 604 speakers, Pak.
Were all having some fun. Get some kicks in before Covid takes us out.
DT667
The plates are not red. Not sure why you keep going after this...maybe you don't understand colours or something simple like that.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
"I go well beyond red there DT... :) If you knew anything about it, you'd say I was being too conservative and should run them hotter."
Your own words , Pak.
Pak claims to "go well beyond" red plating this PP tube finals.
It is probably just the E-Linear final and you are putting some kind of marketing spin on it.
If destroying tubes by over biasing is part of your audio club program, good luck with it.
DT667
Your ignorance is showing again. I can't help you because you don't want me to. It is that simple.
It has nothing to do with the circuit. The tubes are not over-biased either... :)
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
You got nothing for us.
No trade show experience, no profession audio product company experience, not even a project website and trying to take credit for Pete Millet's E-Linear circuit designs.
It's not a good look overall.
But you think you can build a tube amp better than DF and the MLP?
The laugh riot never ends with you, Pork Professor.
Enjoy those red plate sweep tube finals.
DT667
and again your ignorance comes right through. Thanks DT for making it clear. You are just not worth the trouble required to rub your nose in just how badly you have it wrong.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
ask him why it is called E-Linear...
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
"ask him why it is called E-Linear..."
PM invented the E-Linear circuit and Pak tries to take credit with no legal right to do so.
It is called "stolen valor" and also morally wrong, sir.
Expect a letter from Mr. Millet's legal team.
DT667
You have no idea what you are talking about.
He clearly can't be bothered to read the whole thing.
Steps to realize how wrong he is: read the whole thing, esp the part where Pete attributes the naming...and then go ask that principal as to what was going on at the time...
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
you've had almost this exact exchange with him before Douglas
this should be enough for you to conclude that you're being trolled
actually, if not for his rep as a DF fanboy, that poster should be beneath your notice, since all he does is troll as a proxy
anyway, those in the know appreciate your contributions
with regards,
You are no doubt on to something there, and I thank you for the reminder.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
GOOB Sale......one of the MLP boys shorting them.
Edits: 09/21/21
12 watts would ideally be a different kind of amp.
How about a 12BZ7A (using both sections) driving a Lundahl
interstage, driving a pair of 6L6GC-- Sylvania,
or Russian, of course! One can, of course, D.C. couple this amp.
IF you do that, the 12BZ7A is out-- too much gain. Look
for a tube with about 40-45 amp-factor for that.
Run the 6L6's ultralinear-- or use separate G2
voltage that is a bit UNDER the 6L6 plate voltage.
ALTEC-- and myself-- proved that sounds mucho betra!
This G2 voltage must be clean-- a case for several
R.F. chokes and local high-quality (Mundorf
Silver/Oil, etc.) caps.
Alternatively, use a 6SL7 at the front. The interstage tranny
will be a different ratio. Get these ratios from the U.S.
Lundahl Rep-- K & K Audio in North Carolina.
-Dennis-
DF,
Maybe we should "red plate" our SE finals the same way our buddy Pak runs his PP sweep tube amps for max power.
If a tube lasts more than 5 minutes, we are not biasing hard enough.
The marketing slogan could be , "Tubes - better red and dead, than biased for long life."
DT667
But that surprises nobody. Even your accusations are made up.
Now if you'd apply some of that imagination to your audio system design, It could be you might come up with something earth shattering...possibly even something as brilliant as a high gain tube directly coupled to a DH power triode.
I do doubt it though.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
In my imagination a high gain tube is good for guitar amplifiers. The lack of good high frequency bandwith is perfect thereView restrictions here: https://deathtube667tubeamps.com/se-2a3-schematics
Look at the electrolytic capacitor in the signal path of the both tubes.
Edits: 09/27/21
One of the classic effects-box amplifier tricks. I like a heavily loaded pentode as input for a guitar pickup. The low capacitance input is not without value... :) and the low plate load enables more drive as well as the asymmetric, even HD content.
It's entertaining for sure... :)
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
Dennis,
You wrote:
"How about a 12BZ7A (using both sections) driving a Lundahl
interstage,
Alternatively, use a 6SL7 at the front. The interstage tranny
will be a different ratio"
Did you actualy tried this? Do you know that a tube with a high Ri is not a good choice for an interstage?
Do you also know that that 12bz7 is not a very lineair tube?
Which Lundahl you want to use, you know how to use them?
12BZ7A is a lot like a 12BH7A-- in construction.
The 12BZ7 series has more grid turns than the 12BH7--
it's amp-factor is 100.
Both tubes, due to their large plate, grid and
cathode areas, are capable of handling large
momentary current pulses-- which is what music is.
Some would ask, "why not use the 12AX7-- same
amp-factor, but lower noise?" Not enough plate area
for many driver apps-- it works well on efficient
speakers with d.c.-coupled amps-- it can drive the output
tube-- a 2A3-- but cannot go further-- it is smaller in
active area.
I run them all the time on d.c. couples-- it works like
gangbusters on my speaker systems.
The speakers are not super-efficient,
I'm running GPA 604's inside a large MLTL cabinet which is
heavily built-- no unwanted resonances which steal
power.
Lundahls I have some experience with-- Kevin likes to
see 4 ma. or more on most models that would be used for
low power coupling. 10-14 ma. is the accepted published
data on some of those. K & K Audio uses Lundahl interstages
which run 4 ma. on the input side in preamps.
The Lundahls which I did use had something extra
over some other brands-- a large tolerance for
differing operating levels-- in other words, they're
more efficient at low power levels...
An interesting comparison of iron is when we were looking
at various output transformers for the 2A3, and the KT-88
series (KT 120, 150, etc.).
I'm going to stay away from mentioning brands here-- people
can fight over them somewhere else.
We used-- over years-- most of the highest regarded
output trans from Europe, Italy, all the High-End
Japanese brands, etc. We also used several from ultra
expensive custom winders here in the USA and in
Gt. Britain.
Some of these were excellent, and were used in my
commercial amps. Price-- for those amps-- was not
considered due to it being a small part of the overall
cost of building the amp.
Most sounded more than OK at their recommended idle
current levels. a very FEW (all US made) could withstand
LOWER operating D.C. currents than recommended.
While most of them started sounding "Generic"- (not
super high-end) as soon as the output tube plate current
was reduced, a FEW sounded excellent at almost any
lower current level you want to name.
Yeah! A WHOLE lot better design & construction. One
builder, a commonly "generic" supplier, sold us some
of their "super premiums"-- which were designed years
ago by one of their best engineers ever. They had sold
very few of those-- ever.
Guess what? The 14 lb E-I core monster was rated at
over 300 ma. I ran it at 43 ma. on a 2A3. What happened?
It greatly outperformed all the other transformers, period.
In every possible way-- including transparency, speed
and extended highs. Bottom was also better-- expected
because of the large inductance.
Lundahls are like that. They handle ALL of the signal--
for the most part, regardless of idle current.
That's good engineering.
I'm not trying to tell anyone how to run Lundahls.
I have more to learn from them, and frankly don't
yet know how low one can go in ma.. and which
models will allow that. Just as in tubes, they're
going to differ in ways that the specs don't show-- we'll
have to find out-- as usual.
6SL7 can be run in the current range you may need if
you have the right Lundahl. I recommend calling someone who
knows their line better than I do, and getting an
Engineer's ideas-- from several engineers.
If you D.C. couple a 6SL7-- amp-factor is about 70--
you'll have a little extra gain into a 2A3, say over a
triode of 40-45 A.F. This may or may not suit the user--
depends on his system gain. My systems use an amp-factor of 100.
If you choose to run a KT-88, and wish for an industry-standard
amp gain structure, and are D.C. coupling 2 stages,
you'll want an amp-factor of about 33 on the driver.
Don't be afraid of running lower milliamperes on chosen
vacuum tubes. Some are very superior running half or less
of their continuous ratings. Some (transmitting tubes)
must run hotter-- pure Tungsten parts, etc. They bear
no resemblance to receiving tubes, and are used differently.
Running your own plate curves is easy. You need two
variable power supplies, and some milliampere meters,
and your graph paper. Set a plate voltage, set a load,
and insert different grid biasses and plate voltages and
plot all the points on the paper. You'll find out things
about the tube in question that aren't answered on tube
MFR data.
-Dennis-
Dennis,You only speak horse shit. The 12BZ7 is a bad tube and NOT suitable to use with an interstage. The 6SL7 has the same problem but is much more linear
You talking all the time about dc coupled stage but is it really dc coupled? You not using ANY kathode capacitor?
btw, your plate area story is also horsshit
Edits: 09/23/21 09/23/21
"My systems use an amp-factor of 100."Dennis, you use a driver tube with an amp-factor of 100. The stage you are running it in has a gain of less than 70. Edit, after recalculating I should have said "less than 80".
The amp-factor of the tube (also called "mu") is (in theory) the maximum gain you could get from a stage built with that tube and can never be fully obtained. You need a horizontal load line to get you close. 285k ohms will not give you a horizontal load line for a 7b4.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 09/24/21
Correct in that you don't normally
obtain all of a tube's amp-factor.The 7B4-- as a driver for a 2A3 has
been very difficult to beat. Theoretically,
it's not ideal.The trouble is-- it IS almost ideal in
this app.I get a kick out of the 12BZ7A discussions.
That's just THEORY. In actual use, it's
one of the very few driver stage tubes that
actually works!Run your own tube tests whenever you have
a question about whether published plate
curves are correct. In most cases, they're
not even close to the tubes we actually buy and use.-Dennis-
Edits: 09/23/21
Dennis,
You don't understand anything about theory, so don't even try.
Once again you have shown us that you have absolutely no idea how to design.
Please shut up and don't bother anyone with your ideas. It would be so much better if people like you were just put behind a fence somewhere so no one is bothered.
On Sale! And here locally Steve Rouse kept lowering his price in CT until they sold...if they ever did.....would love to know which MLP participant unloaded his?
Edits: 09/23/21 09/23/21
Even if the new price was $3999 it was already $3000 to much. No serious output transformer was used for instance Edcor? Magnequest?
Edits: 09/24/21
What's wrong with Magnequest? My experience with their products was very good.
If one of our posters thinks that Magnequest transformers--
especially the ones I helped design-- are going
to get beat by anything out there-- well, think again.It just doesn't happen-- not musically.
-Dennis-
Edits: 09/24/21
The Magnequest is a great output transformer but it's the wrong impedance (2.5k but should be 3.9k) for the plate voltage and idle current Dennis is running if he cared about HD.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
3.9K is only a calculation, and it is correct
in that published distortion figures will
be lower than 2.5K on a 2A3.
2.5K does something better: at the cost of slightly
higher distortion numbers, it delivers what a 3.9K transformer
(on a 2A3) can't: a tactile feeling-- a direct contact with music.
Only two amp stages, and DC coupling also improve that.
Remember these amps were designed to LISTEN TO, not just
to perform measurements on. There is ZERO chance that a
3.9K transformer will deliver what we have. It will be a
bit cleaner, it will measure better, and 4K has been
tried. Also, 8 ohm speakers have been driven superbly
when connected to the 16 ohm tap of the 2.5K transformer.
The real question is always-- what is the DCR of the power
supply, and are its capacitors small enough and fast enough,
and are the power supply chokes low-DCR and low-inductance
enough so that short-term dynamic peaks are well reproduced..
The 3.9K output trans will handle short-term dynamic events
in a far less convincing manner-- we have experimented with 4K..
2.15K could be better yet. 2.5K is readily available and does
what it should do, musically, so we're very happy with it.
Harmonic distortion is nice to measure and theorize
about. Real life dynamics is something that is essential
to people who enjoy dynamic, complex music.
Speakers are the ultimate challenge here. They must
work, or you go to higher power. If H.F. radiators
have too little surface area, and try to fire from a tiny area,
then all forms of harmonic distortions must be lowered
to control the sub-standard tweeter's departures from
natural dynamic handling of a listening area..
-Dennis-
"Also, 8 ohm speakers have been driven superbly
when connected to the 16 ohm tap of the 2.5K transformer."That would load the 2a3 at 1250 ohms. You are free to like the resultant sound but just because you like it and call it "driven superbly" doesn't mean that those speakers are being driven superbly.
At 1250 ohms the load line for the 2a3 would be rotated towards the vertical and the harmonic distortion will be increased a lot. From 4.44% THD to 11% THD. "Superbly driven" is not a good description of what is really going on here. Distortion is not just a measurement. Distortion can be heard.
I think you like the sound of distortion. Your statement above indicates that you do.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 09/25/21 09/25/21
.
Have Fun and Enjoy the Music
"Still Working the Problem"
Tre,
Remember that Dennis only listens to his amplifiers, the asymetric clipping is part of the sounds: It's amazing, it's the best there is.
Probably he could be presedent too
.
Have Fun and Enjoy the Music
"Still Working the Problem"
Dennis,You are just a salesman trying to sell horse shit like grapes. I have heard salesmen like you many times and I don't believe a word of it.
Edits: 09/24/21
Nobody has ever asked you or anyone else
to believe anything.
What you got is an
honest discussion of why a
thing is done in a certain way.
Designing a system to SOUND LIKE the real thing
is VERY different from getting it to MEASURE like
the real thing.
The SOUND is done IN A ROOM, using human hearing.
The MEASUREMENTS are done with test instruments.
The assessments are made when a recording playback
agrees-- USING THE HUMAN EAR-- with the recorded
venue.
When industry experts HEAR the real thing in a
listening room, they know it and they understand it
when it happens.
No recording engineer who has acquired amps that
were built in the ways that are described-- ever
parted with them. 2 or maybe it's been 3-- audiophiles
have parted with theirs. Audiophiles buy and sell
every day. I have never met an audiophile who kept
a given system together for more than 2 or 3
years-- most change some important part of their
systems every few months.
That's anybody's right-- to do so. No opposition
from me.
But the vast majority of people who have built
systems the right way from the start-- aren't into
parts changing all the time, especially of amps &
speakers.
More than 96% of the amps I ever built in the ways described
in older posts are with the original owners. 100% have never had
a parts failure-- ever.
-Dennis-
And the horseshit continues.....
Dennis, read some of your colleagues' ads . Some of them look exactly like yours.
Never did the trick
Who knows why people like these amps. Sort of like John Hogan who used the same PP Webster Organ outputs for all manner of tubes from 2A3, 300B, 45 etc and people swore by the sound.
Tre',
Given a plate to cathode voltage of 220-250V I'm curious where you come up with 3.9K?
I *think* Paul Joppa may have eluded to successful use of 2.5K here - and 2.5K seems to be a standard value for a 2a3 at 250V.
It seems to me that winding SE transformers with lower impedance requires less skill and engineering for a given design, at least for HF response. From my limited observation.
"2.5K seems to be a standard value for a 2a3 at 250V"
2.5k is the standard for a 2a3 at 250volts plate to cathode but at 60ma.
Dennis runs at 43ma.
"Paul Joppa may have eluded to successful use of 2.5K here"
In fact it is Paul's math I used to determine 3.9k.
Paul's math follows the Western Electric chart very well.
If you want the least HD (across the board) while still having some power, it's a good rule of thumb.
load impedance = (voltage / current) - (2.38 * plate resistance)
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
It is all about the delta V caused by the load R and the delta-i through it. At lower current there is a larger voltage to swing across between idle B+ and where the load line hits 0V. keep in mind this same voltage ideally must be supported by R as you take the idle current towards zero. With too little R, the line is steep and you hit zero current way too soon compared to going the other way... :)
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
thanks tre', appreciate your well explained answer.
Edcor has the CXSE 3K that claims 41H of inductance.. I'd like to give that transformer a try as I hope it may be a good compromise between 2.5K and 5K... (their 5K is rated 50H!)
Edcor dos not say how they measure so you realy don't know what there datasheets mean. This is also true for Magnequest.
They don't give real measurements with a tube for instance (or with a specefic Ra).
If you want realiable measurements Tango is the ONLY manufacturure that dos give accurate measurements.For those who use Lundahl, Tribute or Monolith Magnetics.... Please publish your measurements so people here can learn how these transformers behave in real life. I guess only MM will do a decent job (but there site sucks, but is a lot better then Lundahl or Tribute)
Edits: 09/24/21
There is nothing wrong with the Magnequest if it was a cheaper transformer (maybe a little more expensive than an Edcore). Or there is nothing wrong with an expensive amplifier if they also used a special quality transformer.
The advertisement was originally for a $18000 amplifier (ok, a silly price) but even $4000 is way too expensive if we look at the components used and or the technology used (no real dc-coupled amplifier)For the Magnequest, it is EI laminates, probably M6 grade 0.35mm. That's the quality of 1960. Nowadays we have M4..M3..M2..M0 and even better than that. Also EI laminates are not very good, we have C cores. The thickness of the lamination of standard high quality c-cores is much thinner, so we have less eddy currents in the core.
For specials we also have old-fashioned Permalloy but that too is pre-historic. We now have Amorphous and nanocrystalline cores.And I've been working with amorphous cores for over 30 years so people here are asleep......
Edits: 09/24/21 09/24/21 09/24/21
"There is nothing wrong with the Magnequest if it was a cheaper transformer (maybe a little more expensive than an Edcore). Or there is nothing wrong with an expensive amplifier if they also used a special quality transformer.
The advertisement was originally for a $18000 amplifier (ok, a silly price) but even $4000 is way too expensive if we look at the components used and or the technology used (no real dc-coupled amplifier)
For the Magnequest, it is EI laminates, probably M6 grade 0.35mm. That's the quality of 1960. Nowadays we have M4..M3..M2..M0 and even better than that. Also EI laminates are not very good, we have C cores. The thickness of the lamination of standard high quality c-cores is much thinner, so we have less eddy currents in the core.
For specials we also have old-fashioned Permalloy but that too is pre-historic. We now have Amorphous and nanocrystalline cores.
And I've been working with amorphous cores for over 30 years so people here are asleep......"
I had the Magnequest DS-025. I believe they were M4 lams on that transformer.
It would be nice if you have real data. M4 is better but compared to HiB in some Japanese transformers still lower quality (a lot lower).
Have you read this?
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
"Have you read this?"
Interesting he seems to prefer M19 for SE transformers.
"due to no zero crossing"
"M6 is more costly and its attributes not used in SE service"
Tre' I believe you have some Electraprint SE transformers. Are they M19?
(if my memory is wrong I apologize!)
It's not so important what the people at Electra print think about core materials because they don't understand the datasheets about those materials.Look at:
https://www.nipponsteel.com/product/catalog_download/pdf/D003jec.pdf
and
https://www.nipponsteel.com/product/catalog_download/pdf/D005je.pdfM19: http://lookpolymers.com/pdf/AK-Steel-DI-MAX-M-19-Nonoriented-Electrical-Steel.pdf
For those only interested in the outcome of those data sheets: M19 is even more junk (it's for electro motors)
Edits: 09/30/21 10/01/21
Perhaps Denis is interested in non-oriented SiFe core?
North American Iron is where its at Mr Kobasa.
Hammond, Heyboer, Edcor, Electraprint, Intact. (in no particular order)
Exotic core materials are fine - we have that too. If you can hear the difference. I personally don't have the need for that level of detail.
It helps that we have excellent ganja and beer...so really even the old tubed tabletop radio sounds pretty good after enough substance enhancement.
Of all the names you mention, none make transformer cores, but you can find many good transformer core manufacturers in the US of course.
About the ganja and beer I'm not sure, I bet both are better in Europe (although I don't use ganja but Nederwiet is famous). But I agree that if you use it, none of that transformer stuff matters anymore.
the sharpest guy I know with regards to output transformers is Pakprotector.. he's told me of Thomas and Skinner and their SuperOrthoSil laminations.
I looked to order some in M4 but the quantity was beyond my budget. As we agree - for my tastes and listening conditions I may not benefit from higher quality laminations.
https://www.thomas-skinner.com/transformer-laminations/
I know this company. Skip them...
If you want good prices and good material: China
The best market for good cores is asia in general. All the big manufactures are there (also USA based companies)
I know this article, just skip it because it's not very good.
Got it. I used them at normal 2A3 op points. They sounded pretty good compared to other outputs I had in hand.
I should be beaten mercilessly.
For people who know how to design: fixed bias is very reliable and sounds best (also measures best)
ElBlanco (Thanks!" did us a favor and put up
Robert B. Tomer's "getting the most out
of vacuum tubes".
ElBlanco posted this Sept. 19 on this forum.
I read it all. You might enjoy Tomer's
discussion of fixed bias amps and the
history that led up to them.
Today, commercial fixed bias amps track vacuum
tube conduction individually, and automatically
correct bias to get equal idling currents.
Of course, in the case of unequal tubes (always),
the tubes are forced to idle at the same currents
by applying unequal grid bias. While this
arrangement allows the use of unequal tubes, and
also obtains more consistent power output, one must
also admit the truth:
When the incoming signal sees different bias on
supposedly "balanced" (current-balanced only), tubes, guess
what happens to your signal when it encounters
unequal grid bias on different tubes?
You guessed it: smeared music.
-Dennis-
Quoted:
When the incoming signal sees different bias on
supposedly "balanced" (current-balanced only), tubes, guess
what happens to your signal when it encounters
unequal grid bias on different tubes?
You guessed it: smeared music.
End quote.
I just love how y'all think you understand, and generate plausible sounding justifications, and yet are so completely wrong. Keep right on illustrating this for us Mr. Tube Amp Salesman.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
Salesman is generous. How many do you think he has "sold". I see them
Used occasionally. From Steve who can't seem to unload them.
Edits: 09/21/21
.
Have Fun and Enjoy the Music
"Still Working the Problem"
He likes the "loudness" button on amps. You cam hear the difference when you press it.
I have no idea what BS goes on at trade shows to lead you to believe they matter. One, I sure don't live at a trade show. Two, I have been at this long enough to know that a 3W amp is going to suck more often than it does not( unless I am trying for a tone box...in which case I may design to 2.634 Watts just to make it sound exactly like I want it to).
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
". Two, I have been at this long enough to know that a 3W amp is going to suck more often than it does not"
Douglas,
Given that I know you've got very efficient speakers this statement is interesting for me.
What would you consider a minimum power level given speakers in the 98db+ realm?
I'm pretty satisfied here with 2-4 watts, be it PP or SE...so I guess I really don't know what I am missing!
Look at this datasheets and you know why 2500 is not the best option.https://tubedata.altanatubes.com.br/sheets/127/2/2A3.pdf
Even higher is better to reduce distortion with only a litle power loss
Edits: 09/24/21
God bless you for always bludgeoning with truth, but let me give you
What Dennis dos has nothing to with audio nor music.Btw,
Somebody who writes:
"So far, I've never seen an accurate
tube characteristics chart-- or plate-
curve sets that were anything better than a set
of suggestions, and nothing more.Even those figures vary considerably. A good
example is the 12BZ7A. Some charts look as
if the tube isn't linear enough for audio.Charts from another MFR. looks just like a good 6SN7.
SO__ what gives?"Dos has no clue of datasheets or not have any equipment to check those curves.
Edits: 09/19/21
I am pretty sure that he can improve a lot in his amps, starting with some decent output transpormers
"I am pretty sure that he can improve a lot in his amps, starting with some decent output transpormers"
Better OPT's would helpful.
Or buy one of the latest Class D amplifiers with the new Infineon power IC.
I was young at the start of the CD player and it was a disaster for real audio. I could tell every digital recording from the good analog ones. Only many, many years later came better digital recordings but they could be counted on one hand and they were still worse than the best analog recordings I had.
Still I think digital sounds flat and everything looks alike too. A good friend of mine who has owned a record and CD store for 55 years says: we do like music and then picks up an LP. He owns "Swing Master"Class d ? not even close to real audio, you must be death if you like this.
Edits: 09/19/21 09/19/21 09/19/21
"Still I think digital sounds flat and everything looks alike too."
No way some scratchy 78's played on a Victrola have the same fidelity as today's high resolution audio streams played into a really good Class D amp.
Cleaning vinyl disks has to be one of the least enjoyable aspects of this hobby.
To each his own.
DT667
TK is wanting to know how the makers of the tubes came up with the data in the data sheet. What were their methods, etc.
You pointing out that TK is not doing things the MLP approved way is not helping or in any way answering TK's question.
You sound like a broken record. You just repeat the same thing over and over again even when it's not relevant to the question.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
I do appreciate the fact that a few people offered typical numbers for transformer loss. But then, that information leads directly to what I thought would be a fundamental concept - how did manufacturers test these tubes to arrive at the output numbers? In some cases, they even provide graphs that plot distortion and output power over load impedance, or over anode/screen voltages. I know it's not much in terms of dB, but it's disappointing to build a 35W amplifier, only to see it clip at 30W. What's worse is that the output rating was claimed many years ago, before line voltages increased to 122-125VAC. Because of that, plus the low ESR of modern caps and a few other details of the work, I had actually expected more than the original spec. 40W wouldn't have surprised me, based on voltages and data sheets. The thing is, how can you set a target for a project under these conditions? Shouldn't the data sheets represent real world numbers? I'm stumped.
--------------------------
Buy Chinese. Bury freedom.
A Dynaco Stereo 770 transformer measured 192 Ohm DCR on the primary; secondary is in proportion, so efficiency is 90%. 300 Ohms with 6.6K will also be around 90%. This is copper loss only; core losses are quite small as the flux level is low for most audio frequencies. There are some additional losses at high frequencies, probably small compared to "mismatch" loss due to parasitic inductance and capacitance.
What is the power at 10% distortion?
I should be able to achieve the rated output at levels around 1% or 2%, just like the manufacturers claim. I don't think the amp can be pushed into clipping without immediately exceeding those numbers. Even then, output power won't continue to increase if it's calculated from pk-pk voltage. It's clamped at the clipping level.
--------------------------
Buy Chinese. Bury freedom.
Taking as an example the Dyna A470, the end-to-end DC resistance of the primary is about 200 ohms. Usually the secondary losses are equal to the primary losses, so total about 400 ohms. The reflected load should be 4300 ohms without the wire losses, so net efficiency is 4300/4700 = 91.5%.Obviously the above is for Class A; I'd expect a bit more loss in Class B, or in Class AB at high power, maybe 1.5 times as much. At 1kHz, iron loss should be low since it's far from saturated.
This won't add up to the 30% loss you reported, of course! Must be something more. A couple possibilities:
1) If the tubes are not fully broken in, the cathodes might not be fully formed, limiting available current peaks.
2) If the transformer impedance including DCR is not matched to the spec operating point, that might reduce available power.
Edits: 09/16/21
Paul, the transformers are specified as 6.6K, and that's what I measured on the bench in testing last year. There's a small difference between the 4-8-16 ohm taps. Based on the voltage drop I mentioned earlier, which occurs at roughly 45 mA idle per tube, I believe the primary DCR is 250-300 ohms. These are not as large as ST-70 transformers.
I worked on the amp today to further optimize the feedback and response. At the end, I made another power measurement. With +430V on the CT of the OPT and +350 on the screens (full output, one channel driven), maximum output is about 30W. The tubes are set to idle at 45-50 mA. Current at full output is about 85 mA per tube, measured with a DVM across the 1 ohm cathode resistors.
The original outputs used with this transformer set (PT and OPTs) were 7591. The manufacturer rated the amp at 35W/ch, probably with one channel driven. I changed the outputs to 6L6, then designed a new driver with higher gain. Open loop gain of the driver is about 350, more than twice that of the original 7199. Output headroom at the 6L6 grids is also more than 6dB, relative to the 60V pk-pk required for full output.
Again, this isn't the first time I've encountered this. I'm becoming somewhat skeptical of power tube ratings in terms of the "typical operating conditions" presented in the sheets. It's also not clear to me how that data is acquired by the manufacturers, and exactly what it represents.
--------------------------
Buy Chinese. Bury freedom.
So far, I've never seen an accurate
tube characteristics chart-- or plate-
curve sets that were anything better than a set
of suggestions, and nothing more.
Even those figures vary considerably. A good
example is the 12BZ7A. Some charts look as
if the tube isn't linear enough for audio.
Charts from another MFR. looks just like a good 6SN7.
SO__ what gives?
What gives is WHAT DOES IT ACTUALLY DO?
TWO ways to find out for sure: (1) run your
own plate curves. (2) LISTEN to what it is
doing, using different voltages & currents.
I do both. Because you have to do it.
-Dennis-
I have, more than once(!), found that my gut feel for "close enough" is not always in fact close enough, and I have to start over with more attention to more details.
In the initial post, you drew attention to a specification for 450v/350v plate to cathode, using 6L6GCs into 6000ohmsCT. But you are measuring something like 415v/350v (after subtracting transformer DCR losses at max power) into 6600 ohms. All these little differences can add up fast ...
My two cents.
Yes, small differences can add up, but I wasn't expecting to see the full 50W. The original spec for the amplifier using this iron was 35W RMS output, and that's relative to even lower voltages. The original circuit only indicates +420V on the CTs and +410V on the anodes, and that's at idle. The higher numbers I posted were measured at full output after the supply sags.
In addition to the original specifications, I've searched for other analogous operating conditions. Several manufacturers show 6L6 variants operating in AB1 at lower voltages than this producing more than 35W. The RCA 807 sheet shows +400V anode / +300V screen into a 6.8K load will deliver 36W. The Sylvania 7591 shows +400V anode / +350V screen into 6.6K will deliver 37W. Are these theoretical numbers? Do they include the losses of test transformers? I don't know, but the amp on my bench is operating at higher plate voltages than this, and I expected to see more.
--------------------------
Buy Chinese. Bury freedom.
Here's the chart from Sylvania. My amp clearly clearly falls in-between the two columns that are outlined. But just for the sake of argument, let's say it coincides with the lesser of the two, the one on the left - +400V anodes and +300V screens. According to Sylvania, that should produce 34 watts (nearly identical to the 807, at 36 watts). So why is it, even with higher voltage on the elements than shown in that column, the tubes aren't producing this much output power? Does anyone here know how manufacturers derived these output numbers for their power tubes? Is there a technical text somewhere that describes this rating system?
--------------------------
Buy Chinese. Bury freedom.
RCA tube division production test method
I'll spend some time on that.
--------------------------
Buy Chinese. Bury freedom.
and this book will provide more color.
What sort of Watts are they in the datasheet? Peak watts, rms?
Were your outputs at that % distortion? And at what freq?
Just some questions that crossed my mind.
As far as I know, the numbers are RMS watts. And I haven't had this issue with everything. For example, single-ended 6V6 and 6BQ5 amps perform exactly as they should, even with relatively inexpensive transformers. However, amps I built recently with tubes and voltages similar to those I outlined produced output in the low- to mid-30s, not the 50W shown in the sheets. I take power readings just below clipping at 1kHz. The tubes simply aren't producing what the sheets show, at least not at the output of the transformers.
--------------------------
Buy Chinese. Bury freedom.
A couple of dB of output power loss w.r.t. theoretical maximum output power is not so much and it is hardly audible in terms of SPL.
Omnes feriunt, ultima necat.
Are you using regulated B+ and screen supplies, with no stopper resistors, and accounting for winding dc voltage drop, and swapped in a few new valves to see if their was any variation?
Vintage or new production tubes?
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
The EL34 amp made enough power to meet my requirements, so I didn't go to much trouble regarding troubleshooting. The output tubes were new EH.
The 6L6 amp is on the bench now. I brought it up with brand new 6P3S-Es in both channels, later subbed one channel with a NOS pair of Sylvania 5881. The 5881 performed slightly better, maybe squeaked out two more watts or so. Plate and screen voltages were measured at full output, just below clipping, and they're close to the chart. I also subbed screen voltage from 300V to 400V with a regulated bench supply. No improvement. I don't know the DCR of the anode windings, but the drop at idle isn't large, I think it was about 5 volts. And just to cover all the bases, the driver has much more gain and swing than is needed to drive the 6L6s to full output.
Again, I really think there's a significant loss in the OPTs, and I came to the same conclusion regarding the EL34 amp. Do either of you know how companies like RCA arrive at those output numbers? Did they actually build an amplifier with output transformers? Or did they load the plates in a different way and make voltage/current measurements?
--------------------------
Buy Chinese. Bury freedom.
On loss in output transformers. Let's take a look at Monolith MagneticsThe S9 (Rp=3300 Ohm) has 0.27dB loss and thus an efficiency of 94% (good)
The S11 (rp=5000 Ohms) has 0.50dB loss and an efficiency of 89% (not special)
That is copper loss. The core also has losses and that depends on signal level, frequency, core quality and even winding technique.
Those losses can be measured but are almost never published.The S11 has in my opinion-if we are looking for a high quality audio transformer-a bit high copper losses. Both transformers suffer from low inductance. This is good for lower copper losses, but less good for low frequency distortion.
I have noticed that MM's datasheet is not very good, in my opinion they are not quite right. I infer this from the specified inductance and the low frequency response they specify. Something is not right there. There are definitely errors in there. (small)
Doesn't take away from the fact that they make the effort to publish data and while I think Tango is lord and master for audio transformers and datasheets they do a pretty good job. Lundahl doesn't do a good job and a well-known foodoo brand like Tribute is just bad.
Edits: 09/16/21 09/16/21
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: